Prediction of Yak Weight Based on BP Neural Network Optimized by Genetic Algorithm

AbstractYak weight is an important physiological indicator of plateau yak. The body weight of yak is required for breeding, supplementary feeding, epidemic prevention, and slaughtering. However, due to the large size and strong wild nature of yak, it is difficult to pull it to the weighbridge, resulting in Common weighing methods being time-consuming and labor-intensive and having certain errors. To better predict the weight of yak, in this study, a BP neural network estimation model based on genetic algorithm optimization is proposed to measure the weight of yak. The results showed that there is a significant positive correlation between yak weight and height, body oblique length, chest circumference, and tube circumference. After optimization by genetic algorithm, the root mean square error of yak weight estimation decreased from 0.090 to 0.048, and the error between the predicted value of yak weight and the actual value decreased from 15.4 to 3.1%. Therefore, the algorithm can accurately predict the body weight of yak, and the research results can provide a reference for the estimation of yak body weight in the future.KeywordsBody weight estimationGenetic algorithmBP neural networkEigenvalue selectionPlateau yakData analysis